Changing worn tools is a major concern in planning operations on machining systems. Strategies for replacing tools range from changing each tool as it reaches its projected tool life, to changing all tools when the tool with the shortest life on the machining system is expended. Intermediate strategies involve changing tools in groups. Each of these strategies has two cost components associated with it: (1) the cost of lost production due to machine tool stoppage, and (2) the cost of unused tool life. The best tool grouping strategy minimizes the combined cost of lost production. In this paper we present an approach for finding good tool grouping strategies from inputs that include the tool utilization for a given machining application, and the tooling and machining system costs. A genetic algorithm is used as the underlying optimization paradigm for finding the minimum cost strategy. An example is presented for a part produced on a machining center.